Unlock instant, AI-driven research and patent intelligence for your innovation.

Hyperspectral image waveband selection method based on correlation coefficients

A technology of hyperspectral images and correlation coefficients, applied in instruments, character and pattern recognition, computer parts, etc., can solve problems such as low efficiency and complex algorithms, and achieve the effect of facilitating subsequent processing and application

Inactive Publication Date: 2015-01-07
HOHAI UNIV
View PDF2 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Supervised methods usually require manual input of training samples to make the algorithm conform to the characteristics of actual data to achieve more accurate classification after band selection; unsupervised methods do not use any prior knowledge, so that fully automatic band selection; however, the existing selection methods are complex and inefficient

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Hyperspectral image waveband selection method based on correlation coefficients
  • Hyperspectral image waveband selection method based on correlation coefficients
  • Hyperspectral image waveband selection method based on correlation coefficients

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] The following examples are only used to illustrate the technical solution of the present invention more clearly, but not to limit the protection scope of the present invention.

[0025] A method for selecting a hyperspectral image band based on a correlation coefficient, comprising the following steps:

[0026] Step 1, record the N hyperspectral images to be selected as S 1 ,…S N , the pixel size of all hyperspectral images is p×q; the hyperspectral image S i Arranged into vector a i =[S 1,1 ,…S 1,q ,...,S p,1 ,…S p,q ], where i∈[1,N], N is a positive integer.

[0027] Step two, put a i centralized where mean(a i ) for a i mean of each element, 1 pq is a pq vector whose elements are all 1.

[0028] mean(a i ) = ΣΣS mn / pq

[0029] where S mk for a i An element in , m∈[1,p], n∈[1,q].

[0030] Step three, calculate and The correlation coefficient ρ ij , where j∈[1,N], i≠j.

[0031] ρ ij The formula is:

[0032] ρ ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a hyperspectral image waveband selection method based on correlation coefficients. The method includes the steps that firstly, N hyperspectral images to be selected are recorded as S<1>, ... S<N>, and the hyperspectral image S is arranged into a vector alpha = [S<1, 1>, ..., S<1, q>, ..., S<p, 1>, ... S<p, q>], wherein the pixel sizes of all the hyperspectral images are p * q; secondly, the alpha is centralized, and then the formula shown in the specification is acquired; thirdly, the correlation coefficients rho<ij> of parts shown in the specification are calculated; fourthly, an absolute cross correlation coefficient matrix C<N, N> is calculated; fifthly, candidate element pairs (i, j) when C<ij> is larger than or equal to lambda are selected, and if c<i1> / c<j1> = c<ik> / c<jk> for any k < [1, N] and k = / i, j, one of the ith waveband and the jth waveband can be removed according to the removal principle; sixthly, all the candidate element pairs (i, j) are removed according to the removal principle in the fifth step, and all the finally-left wavebands are finally-selected wavebands. The algorithm is simple and efficient, after waveband selection is conducted according to the method, follow-up processing and application of the hyperspectral images are facilitated, and the method is simple and practical.

Description

technical field [0001] The invention relates to a hyperspectral image band selection method based on a correlation coefficient, and belongs to the technical field of intelligent information processing. Background technique [0002] The main feature of hyperspectral remote sensing is that imaging spectrometers simultaneously acquire information on tens to hundreds of very narrow and continuous spectral segments in the ultraviolet, visible, near-infrared, and mid-infrared regions of the electromagnetic spectrum, obtaining a complete continuous image of each pixel. the spectral curve. While obtaining a higher spectral resolution capability, it also brings about an increase in the amount of data. Moreover, the correlation coefficient between the hyperspectral image bands is often very high, so it contains a lot of redundant information, which will cause a waste of storage and processing capacity. How to remove these redundant information from the original data without losing im...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06K9/66
CPCG06V30/194G06F18/241
Inventor 李昌利高红民张师明徐立中
Owner HOHAI UNIV